P1‐378: longitudinal cognitive and brain volumetric changes to distinguish alzheimer's disease from hiv encephalopathy in those with hiv over age 60 years

An emerging population of HIV-infected elders is at risk for Alzheimer's disease (AD). A critical clinical gap exists in how the well-documented cognitive and brain volumetric effects of HIV can be distinguished from AD, particularly during early AD. We investigated longitudinal neuropsychological performance and brain volumetric changes of HIV-associated neurocognitive disorder (HAND) compared to Mild Cognitive Impairment in HIV-uninfected participants with early AD (MCI/AD). We analyzed selective differences seen when 61 HAND participants were contrasted to 59 demographically-matched controls versus the differences seen when 85 MCI/AD participants were contrasted with 119 demographically-matched controls. We examined neuropsychological testing changes and longitudinal structural brain atrophy rates among regions of interest (ROIs) over approximately 2 years. An averaged standardized global score (NPZ) was calculated at both time points. We performed linear regression models to investigate the association between disease and cognitive change over time (difference between NPZ by disease group) controlling for baseline score, follow-up time, age, sex and education. Linear regression models predicting atrophy rates were adjusted for age and sex. The mean(SD) age was 65(5) for HAND and 73(5) for MCI/AD participants. In the HAND and MCI/AD groups, 55(93%) and 71(60%) were male, respectively. Participants and respective controls were similar in age and sex within disease strata. For HAND participants, mean(SD) nadir and current CD4 counts were 153(139) and 547(243), respectively. All were on antiretroviral treatment with undetectable HIV-RNA. Compared to controls, MCI/AD exhibited significant cognitive decline (p=0.012), whereas HAND showed no decline. However, in adjusted regression models within each diagnosis group, disease vs. control was associated with cognitive decline over time (NPZ(95%CI)) [HAND:-0.33(-0.56/-0.11), p=0.004; MCI/AD:-0.51(-0.87/-0.14), p=0.007]. Both disease groups were associated with faster atrophy rates compared to respective controls (Table 1; Figure 1). Voxel-based analyses highlighted differential increased atrophy rates in select regions associated with MCI/AD vs. HAND (Figure 2). Longitudinal atrophy rates in HAND and MCI/AD compared to controls. Longitudinal atrophy rates, significantly higher in both MCI/AD and HAND groups, identified in the caudate (indigo), putamen (yellow), nucleus accumbens (red), brainstem (green), temporal gray (teal), and left precentral gyrus (magenta). Additional regions exhibiting higher atrophy rates in MCI/AD compared to controls. Longitudinal regional differences between MCI/AD and controls additionally include the lateral ventricles, parietal gray (cyan), parahippocampal gyrus (green), left precuneus (violet), hippocampal gray (red), right precentral gyrus(yellow) and frontal gray (gold). Longitudinal detrimental cognitive changes and accelerated brain atrophy are observed in both disease groups compared to age-matched controls. HAND demonstrated differential focal atrophy in ROIs previously described in HIV, whereas MCI/AD showed widespread increased atrophy. These findings provide early guidance for clinicians in differentiating HIV-related neuropathological processes from Alzheimer's disease.